Because the electromagnetic spectrum is a limited resource, a number of regulatory and technical limitations constrain its utilization in wireless applications, such as communications. Regulatory agencies, such as the Federal Communications Commission (FCC), typically allocate finite slices or bands of the spectrum for various wireless uses, geographic areas, transmission power, and particular operators. (Although this disclosure primarily refers to wireless systems, the descriptions herein may be equally applicable to wired systems, such as those that utilize frequency division multiple access (FDMA).) Also, because certain types of wireless spectrum applications/uses practically operate in a limited range of frequencies, the amount of usable spectrum may even further be constrained by the technical requirements and limitations of a particular application or system architecture. Finally, emerging voice and especially data services generally require increased bandwidth over past services, thereby further increasing the demand for spectrum. As a result of such regulatory and technical constraints on spectral use, as well as the increasing demand for spectrum for wireless services, spectrum allocations are increasingly scarce and expensive. Therefore, there is growing need for methods and systems for increasing the efficient use of such spectrum allocations.
Unfortunately, it is virtually impossible to design practical transmitting devices that completely restrict their transmissions to some assigned band. Transmissions in one band (referred to herein as in-band transmissions) invariably cause some energy to be transmitted into other, especially adjacent, bands (referred to herein as out-of-band transmissions), potentially causing interference to communications in those bands. For this reason, regulatory and technical limits are typically imposed on the maximum allowable power level for transmissions that are out-of-band. (“Band” as used herein refers to one or more contiguous frequency blocks) For this and other reasons, most systems and/or regulations impose not only a predefined maximum power level for in-band transmissions, but also impose a predefined maximum power level for out-of-band transmissions, the latter of which is typically far less than the former, and ideally (but not necessarily in practice) zero. Although these in-band and out-of-band power limits are specified independently, as a practical matter, reductions in in-band transmission levels generally reduce the level of out-of-band transmissions as well.
In some circumstances, for example, as relating to common regulations for cellular voice and/or data systems, predefined maximum power levels for in-band and out-of-band transmissions may further be distinguished for uplink versus downlink transmissions, as shown below in Table 1:
TABLE 1In-BandOut-of-BandMax. PowerMax. PowerDownlink TransmissionPD-IPD-OUplink TransmissionPU-IPU-OAs Table 1 shows, there may be a predefined in-band maximum power levels for downlink transmissions, PD-I, a predefined in-band maximum power level for uplink transmissions, PU-I, a predefined out-of-band maximum power level for downlink transmissions, PD-O, and finally a predefined out-of-band maximum power level for uplink transmissions, PU-O. It should be appreciated that, depending on various system architectures, regulations and other factors, all or a combination of these quantities may be equal or unequal. It should also be noted that, depending on the intended use of the communications system and a regulatory body's use of terminology, the downlink transmission power limits can be phrased instead as limits on “fixed” transmitters (e.g., base stations) while the uplink transmission power limits are phrased instead as limits on “mobile” or “portable” transmitters (e.g., cellular telephones or other remote terminals).
It is common for regulatory bodies and system designers to employ a band allocation scheme that reduces the risk for such adjacent-band interference. For example, in a cellular communication system, two cellular sites that are in close geographical proximity are typically allocated non-adjacent bands to reduce the possibility that the out-of-band transmissions originating from one cell will affect communications in the other cells using an adjacent band (the level of out-of-band transmissions from a transmitting device generally decay with the frequency separation from the device's nominal operating band). Similar non-adjacent allocation of bands for communication devices in geographic proximity are utilized in various other types of wireless applications as well.
Because geographic separation of adjacent-band devices is not always practical or effective by itself to prevent inter-band interference, other techniques sometimes are employed to reduce the effect of out-of-band transmissions and resultant inter-band interference. For example, a transmitter is typically equipped with power amplifier filters to reduce out-of-band transmissions. However, in the case of transmit filtering, even the most complex and expensive filters are not ideal in allowing full power in-band transmission, while completely eliminating out-of-band transmissions. In fact, most practical filter designs suffer from insertion loss (i.e., decreased or irregular power delivery for in-band transmission) and/or some level of out-of-band transmission. Unfortunately, as transmit filters are designed to decrease the insertion loss, the amount of out-of-band transmission is typically increased and/or the filter design becomes expensive. While filters at a receiver can be effective in reducing the effects of powerful out-of-band signals (relative to the receiver's operating band), they can not be used to mitigate interfering signals that fall within the receiver's operating band. Hence the mitigation of interference falling within a receiver's operating band are primarily achieved through the removal of the interfering device in frequency and space, and reductions in the amount of interference that it produces.
A common recourse, often employed in conjunction with transmitter filtering, for mitigating out-of-band transmissions (e.g., to ensure compliance with out-of-band transmitted power limits imposed in some circumstances) is to provide “guard bands” between adjacent spectral bands. This technique is described with reference to FIGS. 1A and 1B.
FIG. 1A illustrates the provisioning of adjacent bands using guard bands, in accordance with the prior art. For illustrative purposes, three bands 102, 104 and 106 are depicted. Each band represents, for example, a contiguous frequency block that facilitates one or more channels that may be available to a particular device or system. A “channel” as used herein refers to any one or combination of conventional communication channels, such as frequency, time or code. For example, band 102 may facilitate channels that a first cellular base station employs for communicating with one or more terminals within its cell, while band 104 may facilitate channels that a second base station that is adjacent to the first base station may employ for communicating with terminals within its cell. Of course, each band may facilitate one or more channels used for communication in a peer-to-peer system.
The bands 102 and 104 border at band edge 108, and similarly, the bands 104 and 106 border at band edge 110. Reference numerals 112, 114, 116, and 118 represent reserved band-edge guards or simply, guard bands, representing portions of bands that, because of their proximity to another band, are not utilized for transmissions in order to prevent adjacent-band interference—that is, out-of-band transmissions affecting users within a band and in-band transmissions affecting users of other bands. The guard band 112, for instance, represents a frequency band within band 102 that is adjacent to band 104 and is therefore reserved; that is, one or more devices that utilize band 102 or a portion thereof are restricted from using one or more channels in guard band 112 in order to prevent interference with devices operating in the adjacent band 104 and also to further minimize the likelihood of significant interference from band 104 devices to band 102 devices. In turn, band 104 includes a guard band 114, which is adjacent to band 102 and is therefore reserved; that is, devices operating in band 104 are restricted from utilizing any channel(s) in guard band 114 for transmission in order to prevent interference with devices operating band 102 and also to further minimize the likelihood of significant interference from band 102 devices to band 104 devices. Similarly, band 104 has a guard band 116 adjacent to band 106, and band 106, in turn, has a guard band 118 adjacent to band 104, to prevent adjacent-band interference between the bands 104 and 106.
FIG. 1B is a graph depicting transmitted power as a function of frequency for in-band and out-of-band transmissions, in accordance with the prior art. As shown, an ideal filter design would provide maximum allowable (as typically defined by a particular system's or regulation's limits) transmit power delivery for in-band transmissions up to a band edge, and provide zero power out-of-band. However, using conventional filter designs that do not provide ideal vertical edge cut-off, systems usually allow some maximum out-of-band power to be transmitted, as shown in FIG. 1B.
In part because of these non-ideal characteristics of filters, as also depicted in FIG. 1B, guard bands are typically included to prevent the effects of out-of-band interference, and the greater the amount and size of the guard bands, typically, the less interference is experienced between bands. On the other hand, because guard bands are essentially not utilized for communication, their use also results in an inefficient utilization of spectrum.
Thus, what is a desired is a method and apparatus for mitigating inter-band interference while improving spectral efficiency over prior art methods.